Fibromyalgia (FM) is a very common rheumatologic condition yet providing an effective treatment for an individual patient remains a challenge. To improve clinical treatment, better understanding of the effectiveness of new drug regimens and the factors effecting patients' responses to these treatments is needed. Anti-depressant medications have been used to treat patients with FM. However, most studies have reported that only about one third of patients show significant improvement with these treatments. A recent study reported that a combination therapy of amitriptyline and fluoxetine (AM + FL) resulted in significantly greater improvement in patients' symptoms as compared with either drug alone. As part of medical practice, physicians and patients often try new, potentially beneficial therapies to assess their effectiveness for the individual. If these studies could be carried out in a scientifically rigorous manner, the collective information could contribute greatly to our understanding of patients' responses to medical treatments. We have developed a method for effectiveness research which uses patient-focused N- of-1 trials and then combines these trials' results to obtain population estimates of treatment effectiveness and to aid in treatment decision-making for an individual patient. This proposal aims to prospectively apply this methodology to compare the effectiveness of the combination therapy AM + FL versus AM alone in the treatment of patients with FM. We propose to carry out N-of-1 trials to compare the effectiveness of AM vs. AM + FL for patients with FM using individual patient (N-of-1) trials. We will analyze the resulting data using the combined N-of-1 methodology to assess overall treatment effectiveness and compare this to results from a prior standard center-based trial. We propose to extend the use of N-of-1 trials into community practices to enable comparison of center-based and practice-based results. Through this broader patient inclusion we will attempt to identify potential patient characteristics which may affect treatment response variation. The results and feedback from both patients and physicians participating in this study will help to develop a framework that will allow transportability of this approach to effectiveness research to the study of other diagnoses and treatments.